Utilizing the Back-EMF

Thread Starter

Trismagistus

Joined Jan 4, 2010
19
Is it very possible to run another low side power mosfet to capture back emf, send it to a capacitor bank and store it there to either boost for acceleration or to send back to the battery to extend the range of the EV? Would this be a regenerative hybrid since we are then using this method full time not just for regen braking? Thanks.
 

GetDeviceInfo

Joined Jun 7, 2009
2,192
This is one of those free energy things. If you extract power that is generated within your motor, you will have to add extra power to make up for what you've extracted, and within that process have efficiency losses. So it's best not to bother in the first place.
 

Ron H

Joined Apr 14, 2005
7,063
Is it very possible to run another low side power mosfet to capture back emf, send it to a capacitor bank and store it there to either boost for acceleration or to send back to the battery to extend the range of the EV? Would this be a regenerative hybrid since we are then using this method full time not just for regen braking? Thanks.
If you're talking about the flyback EMF due to switching an inductor, any energy you capture in a capacitor or secondary battery will be provided by the main battery, so no energy is gained.
 

BillB3857

Joined Feb 28, 2009
2,570
If , on the other hand, you are wanting to use the back EMF for dynamic braking, it is very plausible. During operation, the back EMF is a fundamental element in reducing the line current flowing into the motor. Slow the motor down by loading it, speed drop reduces back EMF, current goes up, torque goes up.
 

Thread Starter

Trismagistus

Joined Jan 4, 2010
19
Well first let me say thank you to the folks trying to help me out. Second, let me get something straight so there's no confusion about what I'm asking for.

I'm looking for a 36 V DC unipolar motor driver circuit that is extremely efficient. I don't need a PIC, just a PWM linked to the Mosfets.Nothing mentioned about "free" energy, and if folks ever find "free" energy then more power to them pun intended.

My idea actually came from a circuit from an MIT EV project that posted the circuit but some of it isn't necessary for me like the tachometer.

My two free Mosfets I'd like to use are the RFG45N06 and the RFG70N06.
 
Last edited:

beenthere

Joined Apr 20, 2004
15,819
How about some parameters to go along with
extremely efficient
One might point out that a switch and two heavy gauge wires are the acme of efficiency. And, even then, the motor's efficiency doesn't get affected.

A PWM circuit is an approach to the switch-and-wires approach. By turning the FET on and off as quickly as may be minimizes the losses.

Parameters, please.
 

beenthere

Joined Apr 20, 2004
15,819
My two free Mosfets I'd like to use are the RFG45N06 and the RFG70N06.
Those are devices, not parameters.

"Free" in what sense? Do you wish to use them because they cost nothing, or is there something about either or both FET's that you think particularly appropriate to the as-yet undefined application?
 

Ron H

Joined Apr 14, 2005
7,063
Well first let me say thank you to the folks trying to help me out. Second, let me get something straight so there's no confusion about what I'm asking for.

I'm looking for a 36 V DC unipolar motor driver circuit that is extremely efficient. I don't need a PIC, just a PWM linked to the Mosfets.Nothing mentioned about "free" energy, and if folks ever find "free" energy then more power to them pun intended.

My idea actually came from a circuit from an MIT EV project that posted the circuit but some of it isn't necessary for me like the tachometer.

My two free Mosfets I'd like to use are the RFG45N06 and the RFG70N06.
I don't see how this relates to your original post.:confused:
 

beenthere

Joined Apr 20, 2004
15,819
Is it very possible to run another low side power mosfet to capture back emf, send it to a capacitor bank and store it there to either boost for acceleration or to send back to the battery to extend the range of the EV
This is a question about making a perpetual motion device.
 

Wendy

Joined Mar 24, 2008
23,415
I don't know the technique, but I know some energy is recovered during breaking on hybrid vehicles. Some, but not all. Basically it extends the mileage a bit.
 

SgtWookie

Joined Jul 17, 2007
22,230
Is it very possible to run another low side power mosfet to capture back emf, send it to a capacitor bank and store it there to either boost for acceleration or to send back to the battery to extend the range of the EV? Would this be a regenerative hybrid since we are then using this method full time not just for regen braking? Thanks.
Sorry it's taken me a while to get to you. There are a lot of people with a lot of projects that are currently "in the works", and there are just not enough hours in the day...

You'd jumped into the middle of another thread that I was pretty deeply involved in, and I'm afraid that I might've come off as being rude. Don't take it that way; you have to understand that "hijacking" of topics makes things really hard on the regulars who are trying to help out.

What I was talking about in the other thread was using a power MOSFET as an "ideal flywheel diode" for the current flowing through the motor; it really had nothing to do with back-EMF. Back-EMF+mechanical friction is what limits the maximum speed of a universal motor.

If you tried to store the energy instead, it would act like a brake to the motor.

However, the synchronous switching of a MOSFET as an ideal flywheel diode is a concept that you might be able to apply in your electric powered skateboard.

Braking is another matter entirely. You can get an effective braking action for a universal motor by shorting out it's terminals. You might also charge a capacitor during that time, which would lessen the braking effect. However, you would need a very large (and expensive) capacitor to store the charge, and you would not get much of a run-time from the stored charge.

The discharge curve of batteries is completely different from the discharge curve of capacitors.
Batteries stay within about 10% of their rated voltage until they are nearly discharged.
Capacitors lose most of their voltage in the first time constant. If you don't know about time constants, please look in our E-books in Vol I under RC Time Constants; the links to our E-books is at the top of every page on this site.

As far as your free MOSFETs; free is a powerful motivation to put them to use. However, we really don't know what your requirements are. The rest of your project might be made more expensive by using "free" MOSFETs, if they don't meet what is required of the circuit.
 

SgtWookie

Joined Jul 17, 2007
22,230
Let me get something straight so there's no confusion about what I'm asking for.
Getting things straight helps - but I'm afraid that you are hijacking your own topic.
I'm looking for a 36 V DC unipolar motor driver circuit that is extremely efficient.
OK, this is confusing. Do you mean a unipolar stepper motor driver circuit? Or a universal-type brushed motor driver circuit? It would help to know more about the specifications of the motor; however I'm afraid that this thread is already doomed unless some kind Moderator decides to sort the wheat from the chaff and re-assembles it into another thread. Not likely to happen; we're running low on moderators lately. :(

I don't need a PIC, just a PWM linked to the MOSFETs. Nothing mentioned about "free" energy, and if folks ever find "free" energy then more power to them pun intended.
There are several other threads going on which are applicable to your situation; including the thread you jumped into. I suggest that you just read for awhile, try to follow along with the situations and experiments, and learn.

Download LTSpice, and run the simulations that've been posted. Change things, and see what happens.

My idea actually came from a circuit from an MIT EV project that posted the circuit but some of it isn't necessary for me like the tachometer.
I'm not exactly stupid, but neither am I an MIT grad. ;)

My two free Mosfets I'd like to use are the RFG45N06 and the RFG70N06.
OK, both of those have a voltage rating that's in the ballpark, and the same with the current rating. However, they have an ENORMOUS total gate charge , and they are both TO-247 packages, which won't be very "n00b friendly" to work with as far as cooling requirements go.

Having the large total gate charge means that they will require a high-powered gate driver, or they will spend a lot of time in the dreaded "linear region" which means they will dissipate LOTS of power as heat.

Starting a horse race with a 3-legged horse generally isn't the best idea.
 

Thread Starter

Trismagistus

Joined Jan 4, 2010
19
The mosfets were free from a discarded electric lawnmower.
I looked up the stats on them and figured they would be ok for the voltage and current Im after. 36V DC and 20-30 amps. So say 1000 watts...and it's for a brushed DC motor.

Wire size shouldnt be a problem for me since I have some surplus from an electric wheelchair. I remember not knowing much about current and wire size when I first started and you are right that these simple things mean so much to a power circuit.

So for an efficient circuit that doesn't guzzle down power what ideas can I use in my set up? Or whats the buzz in the techie world as far as better products and design schemes? Thanks for all of the input.

Thanks also to Wookie, you have a lot of knowledge about electronics. I have some really nice heatsinks on hand too.
 
Last edited:

SgtWookie

Joined Jul 17, 2007
22,230
The MOSFETs were free from a discarded electric lawnmower.
Scrounging stuff is good - if it burns up, it doesn't matter as much, as it was free to begin with right? ;)
I looked up the stats on them and figured they would be ok for the voltage and current Im after. 36V DC and 20-30 amps.
Correct.
So say 1000 watts...and it's for a brushed DC motor.
Well, you have to understand that the MOSFETs themselves can't dissipate anywhere near that much power. You have to be able to turn them on and off very quickly, and when it's time for them to be OFF, you have to shunt current around them quickly.

Wire size shouldnt be a problem for me since I have some surplus from an electric wheelchair. I remember not knowing much about current and wire size when I first started and you are right that these simple things mean so much to a power circuit.
There are many resources on the Internet for gross calculations. However, realize that at high frequencies, stranded wire beats solid wire.

On the other hand, flat conductors beat round conductors.

So for an efficient circuit that doesn't guzzle down power what ideas can I use in my set up?
We just don't know enough about your project. :(

I did some Google inquires about your MOSFETs, and found their datasheets. (you can do as I did.)

I don't know what you plan to use for a motor, or it's current requirements.

Hey, this is your project. We can help, but you have to do much of the "legwork" - we won't do the whole thing for you.

If we did, you would not learn anything.
 

Thread Starter

Trismagistus

Joined Jan 4, 2010
19
Here's a circuit I was looking at from some MIT guys http://web.mit.edu/first/kart/.

Theirs is an H bridge design with a few more details than most motor control circuits I've seen. Remember I just need a one directional driver for a brushed permanent magnet DC motor. Like the capacitor bank booster. How cool is that? To my untrained eye it looks like they aren't using a flyback diode because they have a second mosfet in its place which pulses current to a cap bank on the motors off-time? The concept of retrieving back emf pulses of current from a motors off time,for example one that's on a 50% duty cycle, and storing that charge in a cap bank for whatever purpose seems doable to me but I'm not schooled on the subject and have a lot to learn . Doesn't a flyback diode just waste that energy so it wouldn't be utilized?

Wookie you bring up a good point about switching these mosfets ultrafast and shunting them. Since Im building this project around my free parts, hence free energy...lol, I guess I will need to build my PWM to a frequency the mosfets need to switch at?Or do I need a separate gate driver to power and switch the mosfet after the pwm?
 
Last edited:

SgtWookie

Joined Jul 17, 2007
22,230
I received your E-mail.

Sorry it's taking a while to get back to you, but I'm pretty deep in several other people's projects at the moment; including RobbJohnson's project, which is directly related to your own. Consider that what I'm doing on Robb's project will be applicable to what you're attempting to do.

You mentioned this PWM driver: http://www.allaboutcircuits.com/vol_6/chpt_6/9.html
in your E-mail; however it is woefully inadequate for your purpose.

It is very necessary to discuss wire gauge and lengths, gate charges, and switching times in a high-power project - as they are all part of the equation. If you leave something out, you will more likely than not wind up with a loud bang, a bright flash, and lots of smoke.

If building a successful PWM driver for a large motor were easy, Robb would have been finished with it months ago.

Your drivers will require an additional level of complexity simply due to the higher voltage you're working with. You can't exceed the limitations of the MOSFET specifications unless you want to produce loud bangs and smoke.

You're adding another level of complexity, because you want to recover the back-EMF when you apply the brakes. How practical this turns out to be is in question; a skateboard is a great deal smaller than a go-kart.

Meanwhile, I've put in my 18 hours today.
 

SgtWookie

Joined Jul 17, 2007
22,230
So, tell us about this motor. It's what - 1,000W or 1,200W? Is it rated that at 36V, or did you decide to use 36v based on another application?

It's important to know, because that will make a big difference in the actual power. For example, if it was designed to run at, say, 24v @1,200W, increasing the voltage to 36 would result in it drawing a lot more current.

You're going to have to make a decision between efficiency and noise. Since the gate charges are so large with your MOSFETs, operating them at moderate frequencies (like, 40kHz) will require a good bit of power all on it's own.

The lower the frequency, the less power is dissipated by charging and discharging the MOSFET gates. However, if you're in the audio range, you'll hear the motor "sing" due to the pulsing current. How much it will "sing" depends on the motor itself, and the power being run through it.

As far as a gate driver, the HCPL-3120 looks to be the ticket. You'll eliminate most of the kinds of problems that Robb has been running into on the other thread.

If it's going to be a low-speed PWM driver, then you could use a simple 555 timer circuit to switch the emitter of the HCPL-3120.

A remaining problem is powering the circuit. Are you planning on using three 12.6v sealed lead-acid batteries in series? Or do you have some kind of battery pack in mind?

As far as the back-EMF capture goes - that'll be rather complex. It will require either a MOSFET H-bridge dedicated to it along with a bulky capacitor (or a number of them) to store the charge, and a method to either switch it/them in series with the batteries momentarily, or to discharge it separately. It would of course require circuitry to determine when the cap(s) were close to being discharged, as to not degrade performance (or become destroyed due to charging in the wrong polarity) by being in-circuit too long.

There are also reliability and safety ramifications due to the added complexity of designing and building such a circuit. I don't know what your level of skill is, but adding such a feature would really complicate things.
 
Top